Epiretinal implants can restore low-quality visual sensations by artificially inducing activity in the retina of blind patients, but a key limitation remains: stimulation often co-activates passing nerve fibers alongside target cells, producing streak-like percepts and limiting image sharpness. We investigate a new approach to avoid this: very-short electrical pulses of about 10 microseconds duration. Previous work shows that such short pulses reliably trigger responses and increase activation thresholds more in nerve fibers than in cell bodies, opening a window for precise, focal soma stimulation without co-activating fibers.

 

Electrical stimulation of a single retinal ganglion cell at the soma and the axon.

© Adapted by Paul Werginz 2026 | from 2025, IEEE TNSRE

To enhance this selectivity, we combine mouse retina electrophysiology experiments with computer simulations. Our goal is to arrange electrodes so that the electric field is directed as vertically as possible across the cell body. Based on our findings so far, this vertical field profile is key for selective cell body activation. In parallel, we examine return-electrode placements that drive current through the retina, aiming to improve focus and reduce required currents to facilitate translation to implants.

The work plan measures thresholds across different cell types, tests pulse trains up to 300 pulses per second, and maps how single cells respond to electrode position. Simulation models guide practical electrode layouts that are then verified in the lab, creating a fast prediction–test loop. The project is conducted at TU Wien; an international partner supports electrode design and provides custom microelectrodes if needed.

 

 

Potential findings of this project are not specific to retinal ganglion cells as all nerve cells follow the same all-or-none principle. Therefore, our findings will have broad implications on neural engineering and our general understanding of neuronal signal generation and transmission.